> 1. SO I have done the strip size to 2048 which gave me a performance
> boost from 30mb/s to 104mb/s on a bs=1M
>
I find that bigger stripe sizes (than the default 64k) give you worse
write performance and higher iowait but better read performance on
large files.

>
> 2. What would be a difference between these few dd commands, Do those
> simulate different file sizes ? or ?
> dd if=/dev/zero of=/home/lucas/bigfile bs=64k count=8192
> dd if=/dev/zero of=/home/lucas/bigfile bs=128k count=8192
> dd if=/dev/zero of=/home/lucas/bigfile bs=256k count=8192
> dd if=/dev/zero of=/home/lucas/bigfile bs=512k count=8192
> dd if=/dev/zero of=/home/lucas/bigfile bs=1024k count=8192
> dd if=/dev/zero of=/home/lucas/bigfile bs=2048k count=8192
> dd if=/dev/zero of=/home/lucas/bigfile bs=4096k count=8192
>
They simulate different write sizes and also different amount of bytes
written. Multiply the numbers for the bytes written. 64k * 8192

>
> 3. is the stripe_cache_size in bytes? kb? mb?  and what is
> stripe_cache_active tells me?
The # of blocks that currently in use of the stripe_cache.
stripe_cache_size is the # of blocks total.

>
> 4. I noticed that iowait is now at about 40% constant...is that good?
>
No. This is not good.

> What exactly iowait does for me?
>
Your system waiting to write.

John

------------------------------------------------------------------------------
SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada.
The future of the web can't happen without you.  Join us at MIX09 to help
pave the way to the Next Web now. Learn more and register at
http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/
_______________________________________________
Bacula-users mailing list
Bacula-users AT lists.sourceforge DOT net
https://lists.sourceforge.net/lists/listinfo/bacula-users